BEVERS: A General, Simple, and Performant Framework for Automatic Fact Verification

Mitchell DeHaven, Stephen Scott


Abstract
Automatic fact verification has become an increasingly popular topic in recent years and among datasets the Fact Extraction and VERification (FEVER) dataset is one of the most popular. In this work we present BEVERS, a tuned baseline system for the FEVER dataset. Our pipeline uses standard approaches for document retrieval, sentence selection, and final claim classification, however, we spend considerable effort ensuring optimal performance for each component. The results are that BEVERS achieves the highest FEVER score and label accuracy among all systems, published or unpublished. We also apply this pipeline to another fact verification dataset, Scifact, and achieve the highest label accuracy among all systems on that dataset as well. We also make our full code available.
Anthology ID:
2023.fever-1.6
Volume:
Proceedings of the Sixth Fact Extraction and VERification Workshop (FEVER)
Month:
May
Year:
2023
Address:
Dubrovnik, Croatia
Editors:
Mubashara Akhtar, Rami Aly, Christos Christodoulopoulos, Oana Cocarascu, Zhijiang Guo, Arpit Mittal, Michael Schlichtkrull, James Thorne, Andreas Vlachos
Venue:
FEVER
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
58–65
Language:
URL:
https://aclanthology.org/2023.fever-1.6
DOI:
10.18653/v1/2023.fever-1.6
Bibkey:
Cite (ACL):
Mitchell DeHaven and Stephen Scott. 2023. BEVERS: A General, Simple, and Performant Framework for Automatic Fact Verification. In Proceedings of the Sixth Fact Extraction and VERification Workshop (FEVER), pages 58–65, Dubrovnik, Croatia. Association for Computational Linguistics.
Cite (Informal):
BEVERS: A General, Simple, and Performant Framework for Automatic Fact Verification (DeHaven & Scott, FEVER 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-3/2023.fever-1.6.pdf
Video:
 https://preview.aclanthology.org/nschneid-patch-3/2023.fever-1.6.mp4